iEnhancer-RD

iEnhancer-RD identifies enhancers within DNA sequences and classifies them by regulatory strength using RKPK feature extraction and deep neural network classifiers.

Key Features:

• RKPK Feature Extraction Method: RKPK integrates three distinct feature methods to capture the essential characteristics of enhancers.
• Recursive Feature Elimination: The RKPK method utilizes a recursive feature elimination algorithm for effective feature selection.
• Deep Neural Network Classifier: A two-layer deep neural network architecture performs Layer I enhancer versus non-enhancer DNA sequence classification and Layer II strong versus weak enhancer classification.
• Performance Metrics: Independent dataset testing reports 78.8% accuracy in Layer I and 70.5% accuracy in Layer II and indicates performance superior to many existing methods.
• Implementation: The architecture is implemented in Python.

Scientific Applications:

• Genomics research: Rapid identification and classification of enhancers to support genome-wide regulatory element annotation.
• Gene regulation analysis: Mapping enhancers to facilitate understanding of gene regulation mechanisms.
• Disease-associated regulatory studies: Studying genetic diseases linked to regulatory element dysfunction by identifying enhancer elements.
• Personalized medicine and target identification: Identifying potential regulatory therapeutic targets relevant to personalized medicine.

Methodology:

RKPK integrates three distinct feature methods and applies recursive feature elimination for feature selection; a two-layer deep neural network is used where Layer I discriminates enhancers from non-enhancer DNA sequences and Layer II classifies enhancers as strong or weak; performance is evaluated on independent datasets; implemented in Python.